| Literature DB >> 34491747 |
Joel Graef1, Christiane Ehrt1, Konrad Diedrich1, Martin Poppinga1, Norbert Ritter1, Matthias Rarey1.
Abstract
The ever-growing number of protein-ligand complex structures can give fundamental insights into protein functions and protein-ligand interactions, especially in the field of protein kinase research. The number of tools to mine this data for individually defined structural motifs is restricted due to the challenging task of developing efficient index structures for 3D data in relational databases. Herein we present GeoMine, a database system with web front-end mining of more than 900 000 binding sites. It enables database searches for geometric (interaction) patterns in protein-ligand interfaces by, for example, textual, numerical, substructure, similarity, and 3D searches. GeoMine processes reasonably selective user-defined queries within minutes. We demonstrate its usability for advancing protein kinase research with a special emphasis on unusual interactions, their use in designing selective kinase inhibitors, and the analysis of reactive cysteine residues that are amenable to covalent kinase inhibitors. GeoMine is freely available as part of our modeling support server at https://proteins.plus.Entities:
Mesh:
Substances:
Year: 2021 PMID: 34491747 DOI: 10.1021/acs.jmedchem.1c01046
Source DB: PubMed Journal: J Med Chem ISSN: 0022-2623 Impact factor: 7.446